A Scalable High-Performance Priority Encoder Using 1D-Array to 2D-Array Conversion

TL;DR

This paper presents a scalable, high-performance priority encoder architecture that converts 1D arrays to 2D arrays, achieving higher operating frequencies for large bit-widths through optimized design and look-ahead signals.

Contribution

It introduces an optimized (M, N) pair and look-ahead signal approach for scalable large-sized priority encoders, significantly improving performance over existing designs.

Findings

Achieves up to 649 MHz at 64 bits

Operates at 520 MHz for 256 bits

Reaches 370 MHz at 2048 bits

Abstract

In our prior study of an L-bit priority encoder (PE), a so-called one-directional-array to two-directional-array conversion method is deployed to turn an L-bit input data into an MxN-bit matrix. Following this, an N-bit PE and an M-bit PE are employed to obtain a row index and column index. From those, the highest priority bit of L-bit input data is achieved. This brief extends our previous work to construct a scalable architecture of high-performance large-sized PEs. An optimum pair of (M, N) and look-ahead signal are proposed to improve the overall PE performance significantly. The evaluation is achieved by implementing a variety of PEs whose L varies from 4-bit to 4096-bit in 180-nm CMOS technology. According to post-place-and-route simulation results, at PE size of 64 bits, 256 bits, and 2048 bits the operating frequencies reach 649 MHz, 520 MHz, and 370 MHz, which are 1.2 times, 1.5 times, and 1.4 times, as high as state-of-the-art ones.